The prevalence of marijuana use among older adults is rising dramatically. Marijuana is being promoted as a harmless ?herbal? treatment that can be used to treat insomnia, stress, caregiver fatigue, and pain, conditions that are common among the elderly. However, there is a complete lack of data on the effects of marijuana and its active component, tetrahydrocannabinol (THC), among older adults. In trials examining the effect of THC-based pharmaceuticals among younger populations and in specific conditions, serious adverse events reported included dizziness, nausea, somnolence, vomiting, disorientation, confusion, loss of balance, and hallucination. These adverse events of THC-based pharmaceuticals provide a framework to understand the potential harms of plant-based marijuana use in older adults. We hypothesize that marijuana use in older adults is associated with fall-related injury, any injury, and hospitalization for altered mental status and mood disorders. In addition, as smoking is the predominant form of use, respiratory complications are also likely. To address the extensive gap in the literature on the potential harms of marijuana in older adults, we propose to construct a cohort of patients 65 years and older in the Veterans Health Administration (VA) using merged national VA and Medicare data. Specifically, we propose to use validated text processing methods to review medical record notes and identify mentions of marijuana use to preliminarily categorize patients into marijuana users and non-users. We will then perform an in-depth 25-minute telephone health interview with each patient to ascertain exposure history and baseline characteristics. After adjusting for the baseline differences between users and non-users using the comprehensive interview data collected and the national VA data available, we will compare the rates of events among 2250 current marijuana users and 2250 non- users during the follow-up period. Our proposal will accomplish four goals: (1): To use a prospective cohort design to examine the association between marijuana use and the primary adverse outcome of all-cause hospitalization and the secondary composite outcome of all-cause hospitalization and death. (2) To use a prospective cohort design to examine the association between marijuana and the secondary outcomes of fall-related injury, any injury, altered mental status, mood disorders, and respiratory infections. (3) To examine whether specific subgroups of older patients are more likely to experience adverse outcomes. (4) To determine whether there is a level of safe use of marijuana in older adults. Though very little is known about the physical harms of marijuana in older adults, it is likely that use of marijuana is associated with adverse effects. Given increasing use among older adults, a better understanding of the risks of marijuana use in this population is urgently needed to disseminate accurate information to the public.